seo-images-media
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external HTML content from web pages to perform audits and generate image descriptions. This surface is vulnerable to indirect prompt injection, where a malicious website could embed instructions designed to manipulate the agent. \n
- Ingestion points: The skill processes rendered_dom and raw_html in SKILL.md. \n
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the ingested HTML as untrusted data. \n
- Capability inventory: The skill allows usage of Bash and WebFetch, which could be abused if an injection is successful. \n
- Sanitization: The skill does not specify any sanitization or filtering for the external HTML content. \n- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Node.js script as part of its verification logic. \n
- Evidence: node scripts/parse-html.mjs --url is specified in the verification section of SKILL.md.
Audit Metadata